My sim consists of a Logitech G923 wheel and pedals, gear lever and simple desk and gaming seat that I have to secure to the wardrobe with a piano stool and a guitar speaker for it not to move (it still does a little when I push the brake). I suppose it could fall a little short of F1 teams' expectations...
@@docmain999griptape under a diningchair was my setup for years, although a cockpit and better wheelbase and all that sort of stuff adds alot to your quality of life.. it doesnt make you a quicker driver
My room mate blows next to me when there's wind and spits when it's raining. He also kicks me depending on the crash. Much more real phyisics than those "top" simulators.
@unklesuga1644 thats why I am saying that. Cadillac knew for awhile they were gonna be 11th team. As soon as Micheal Andretti stepped down that was what F1 wanted.
@blade0667 i do actually forget about their team🤣 you're right. IF Ferrari is using that Simulator and Ferrari are supplying Cadillac's engines it makes sense
I m very perplexed because ALL teams missed the ‘porpoising’ effect and coming from a mech eng with 40 years exp. in CFD / STR analysis because the repeated bouncing up and down should be simple to solve I think !
Several of the Apollo astronauts were Marine or Navy pilots so they were already trained for high stress environments but of course still needed to practice working the specific controls for the Apollo flight plan so plenty of simulators of all kinds were created, truly amazing for 1960s and 70s 😎
@@philippemichelvidori7248 it's not really just bouncing up and down mate! It's a rather complex problem that I'm not exactly competent to explain to you having only a basic understanding, but what I can tell you is that fluid dynamics get complicated really quickly and when you have to consider flow on all sides, in the venturi tunnel, around the wing, etc. and on top of that distinguish between turbulent and laminar flow... it gets utterly mad.
@@philippemichelvidori7248 The problem with porpoising is that the windtunnel models are static and can't move up or down with regards to downforce, it's specifically to prevent the model from hitting the belt they run on as it would likely explode and create massive damage to both the wind tunnel and the model. So they're stuck relying on CFD simulations which are very performance intensive, you can either run a very accurate simulation on a small section of the car or run a more complete but inaccurate simulation on the whole car, or somewhere in between but never fully both. Porpoising is also a very complex feedback loop making it especially hard to simulate accurately even without the issues i mentioned before.
@philippemichelvidori7248 porpoising IS easy to resolve, you just run the car stiffer and higher. But if you also want to be fast, you have a particularly complex problem.
You can't 1-1 simulate the g-forces the inner ear feels in a sustained way, but you can do it to some parts of the body. Eg your skin won't know the difference between being pressed into the side or your seat in a corner vs your seat pressing into your skin in a corner. Some sim rigs use this trick to simulate the sustained Gs with movable panels in the seat. Don't think it's 100%, but it does sound like 1 more level of immersion where you could quite literally be driving by the seat of your pants ;)
You can! Just make the sim rig actually move, and to sustain the g force, it should move miles away with wheels... Ah crap I just described actually driving the racecar
@@fookapi that's with compressed air, right? not sure if that would have the response time they are looking for. but yeah, there are other ways to simulate Gs without moving the head about
@@alex8373 I’m imaging a full body suit that can apply pressure to different parts of the body to simulate forces. So unlike a traditional g-suit (which I wore as a military pilot), it shouldn’t just compress the legs. Just an idea and yes, it probably wouldn’t be fast enough.
Interesting video, but I would have liked it go way more into detail. “Simple” things like how they manage to keep latency low, make use of synchronization between visuals and movement of the platform, perhaps even sound (you didn’t mention that at all). But more importantly, you sidestepped *how* teams translate a part that only exists on the digital drawing boards to be “fitted” onto the virtual car and then influences the way the simulator behaves. To me, that is the most interesting aspect of this workflow.
Do these simulators have custom software for graphics and track specs? Did they laser scan all the tracks themselves and create their own sim software or are they using 3rd party data? Super curious!
@ that’s what I’m wondering. Like if you’re 2mill in a sim. Is that “enough money spent, let’s use iRacing”. Or is it, “only our own personal laser scans of tracks meets the standards for our super expensive sim”
@@naplesbluesrt i do have a proper rig but i can't use it as i kept stubbing my toe when i got out of bed due to how close my desk is to my bed and just how big the rig is lol
The other team is McLaren. The names of Ferrari and McLaren show up in the SEO results in the short description of SAE International's website when they covered Dynisma as a startup in 2022. I wonder if McLaren decided later to keep their supplier info private, while the search engines still have that information and the webpage is still up.
Sustained g-force is possbile. AMST Desdemona offers 3 g sustained - being a centrifuge. It probably can't achieve the necesary latencies but all I want to say is, you can simulate sustained g-forces.
I wonder if you could simulate gforces in a big wheel with a rotating capsule, if you turn it a bit upside down and accelerate a little over 1G you could counter act the earths g force. The question would be how hard is it to increase Gs fast, can you go from 1 to 6 in fractions of a second. Maybe with a clutch or moving the capsule outward on tge wheel, or both.
My sim consists of a camping table as the desk, with my G29 wheel secured with a chopping block to keep it in place. For my pedals it’s under the camping table for blocks of wood so it doesn’t slide around on my carpet under braking. My gear shifter is on a container with every Harry Potter book staked on eachother as the stand with a flat wood plank and the top with the gear shifter secured to that. It also has wood counterweight blocks.
That was my idea, then the car can rotate 360 ( if needed) to simulate g force. Then the car itself can be on rails itself and go up down the arm spinning, like a camera zoom steeper motors. To simulate another g force potential. While it can also increase / decrease g force as needed ( physics). While the main spinning motor can be slowed and speed up to also simulate more. The guy said you only need to get a percentage of the way there, so 2.5 g feels like 6 in sim ? 😮... I want to try it please !!
As a professional in simracing I've tested and installed many systems. The best of all is the one I still use since 12 years, it also have such frequency filtering no latency, it is affordable you can see on my channel. In combinaison with dlp projectors it has real 5ms latency from my hands to the feel and picture I never feel that alive feeling in any other motion thing
@@Supersonic_racing Pressure on the skin is one thing. Feeling acceleration in the inner ear is another important thing - which can't be influenced by pressure suits.
Really good video - explains all the fundamentals at play in the best motorsport simulators. 3 to 5 milliseconds latency - this is very good, but not the best available in F1 - agree totally that it is one of the critical things, but this sim is not the best for this particular metric. Am certain that they are working to improve this further. Driver reaction time is about 20 milliseconds or so, thus the latency is excellent compared to older driver in the loop simulators. Agree that the slight remaining latency is the main cause of drivers feeling sick - the world normally reacts faster in a real race car. Some really good and honest info about how the lower G-force is used to give a driver the most realistic feeling one can have. Loved it!
Fantastic video! Masterly created and on a very interesting topic at the heart of F1 development. My F1 sim consists of a MX5 and while it also does not achieve the same G-forces it does replicate without any delay many of the minute and sometimes major feelings like the rear end rotating mid corner and stepping out.
I always think that the best way a simulator could traduce g forces could be that the whole unit would be like a ball, where the pilot gets in, and inside of it , it would have like a cockpit, and with enough distance, rounded screens like the sphere in las vegas, but small size. Then, that ball-like unit would have the ability to rotate in any direction with a mechanic structure in the outside, but when youre sitting in the cockpit, with the seatbelts on, the horizon in the screens would be alingned properly in your field of view, so, when breaking, the ball viewed from the outside would rotate forward in the range of motion between 0 and 90 degrees depending on the strengh of the break, but viewed from the inside the pilot would have the screens with him so he wouldnt notice he is inclined towards the floor but he would only feel the weight of gravity as g forces on the break. And that same idea for the lateral g forces they feel when turning on a corner, or to feel speed G forces when accelerating inclining the ball backwards having the pilot facing the sky (without him noticing cause hes inside the screens sphere). Instead of having a big screen attached to the floor and even if you have a simulated cockpit that moves and has some forward or lateral inclinations it wouldnt ever be able to incline much, without it loosing realism because the screen is separate from it and it would break the feeling of a real car-view behavior. I got this idea from a f-16 or f-18 flight simulator I tried once, but it was different than this idea, it would be like 2 cockpits attached in front of the other, then you would have like an axis between them, so youwould get into the cockpit and you would only see like an inside of an airplane cockpit and like a screen where you would see like the airplane view or something like that. And then it would "take off" and you would begin to feel the G force of a real f-16 but viewed from the outside, the 2 cockpits would be spinning like a fan or a blender, and the cockpit would be on each blenders blade looking towards the axis. That way you have the g forces only in one direction, you would only be able to feel "speed" G forces. Not breaking or lateral G forces. Thats why I thought about the other Idea. I dont know much about simulators but it has always been interesting to me that Ive never seen one with the same idea I have in mind, not even in f1 simulators. Maybe thats a bad idea but it makes so much sense to me that it would be way more realistic for the pilot.
You'd still be limited to roughly 1G, much less than the forces in a real car, and you might get some unwanted effects or cues to the driver due to having to rotate forwards, backwards or sideways to apply the g forces
Not able to apply high enough forces, will move the pilot in a manner inconsistent with the vehicle, and a hugely complex video problem (to create something less real).
great video. i really wish home sims like iracing would allow us to render only the elements that we need like in 1:16. if i could only render the outside world and hide the dash i would be a very happy camper
I wonder if parallelization could be used to pre-compute all the predicted values that can happen on the next frame or frames... I mean, if the value of something is '5' in this frame it won't be 50 000 the next frame. It will be 6 or 8. Add 16k predictions matrix and you can pick the solution from a storage. Kind of what our brains do all the time... Computing something takes time but if we can solve all possible predictions of what is going to happen next then you only need to execute.
I'm a bit confused. For the inner ear breaking and pitching forward feels exactly the same. Can't you just simulate g forces by pitching/rolling? Sure you can only simulate 1g of cornering/braking acceleration, but this could be compensatet for with additional weight the driver wears. Instead of the huge screen you would obviously need vr googles or somthing comparabel to fool the driver into experincing only the g forces without a relation to their surrounding. The only downside would be the constant weight pulling driver down. But this way you could have sustained g forces throughout longer corners/braking sections. This exact same principal is used for flight simulators. Maybe I'm just overseeing something, but this would be a logical way for me simulate sustained g forces. Maybe you can help me see my flaw in logic here?
The inner ear knows nothing about the additional weight on the body. For the ear, 1 g is still the maximum sustained force by pitching - leaving 0 g felt towards the butt. If you want sustained g-forces even above 1g that the ear can sense, you need a centrifuge (see AMST Desdemona for example).
The consumer/console games don’t do a good job of decoupling the front and rear axles, but I’d imagine from the remarks about the feel of the tear end under braking, these are beyond an order of magnitude better in that regard. So cool. Would like to know if this company is supplying the NASCAR teams as well.
Hey scott can i ask you something? is the spanish track an AI? because its sound just like your voice, if is a real person well congrats because its so good
I kinda expected the bigger rigs to use some sort of 6 axis system to use gravity to simulate G-force, though i suppose when the cars can pull 6G maybe that is less useful.
Interesting that platform pitch and roll aren't used as surrogates for acceleration as they are in flight simulators on Stewart platforms. One reason is because the visual is not on the platform and would have little screen space left to compensate for the tilt, particularly in roll, and another is that these types of cues produce undesirable rotation rates in the inner ear, ever so more with the required onset of the acceleration, leading to motion sickness. With flight sims, braking is the worst offender, and difficult to tune the filters to achieving a compromise. Surprised 3ms latency (and they probably meant throughput) is required, as the human animal has a mean 1/3 sec time constant - I would have thought 30ms would suffice in these high gain applications. Nice work.
Does anyone know if the Iracing guys took their model to develop software for the race team sims. I can’t remember where I heard that. So a lot of these are based off the original Iracing models
So they derive the acceleration once more to get jolt, that is the rate of change of acceleration. Using a tilting sim would allow for simulating a weak but rather constant centripetal force, but it would produce some other unwanted forces too. It's better suited for aviation sims, where constant g-loading builds up rather slowly, so the required reaction speed is much slower, apart from the little bumps you'd feel from aerodynamic turbulence.
The rate of change of acceleration is Jerk, not jolt. And the rate of change of jerk is snap, then crackle and then pop. Don't say physicists don't have a sense of humour
@@PaulHickford Whatever order of magnitude we look at, at some point we reach a degree of amalgamation where bones behave like play-doh, planets behave like liquids and space-time behaves like a fabric of information. Isn't the universe fascinating! - some ER surgeon pondering about his last job, probably
I found the part where the gentleman explained the “high pass filter” interesting. They use the “jerk” (technical term) to impart the feeling of changes in acceleration and deceleration. The drivers’ imagination does the rest! Cool.
@@stevenhaggar132 Been giving it a bit of thought lately. I reckon three wires, two diagonally out the front and one going straight back. I would mount that to a bearing mounted ring so you can turn your head and mount the ring to a helmet with velcro. I figure you want a system that you can break free off in case of a fault in the system, no responsibility taken for broken necks or whiplash... :) Edit: Scratch the whole ring and bearing idea. Just mount the wires to the top and centre of the helmet, that should allow enough freedom to turn.
@@stevenhaggar132 Thanks for the vote of confidence, I emailed Dynisma and passed on the idea about 10 minutes ago.. Not sure if we'll ever see this for consumer grade simulators but if you are spending mega bucks it wouldn't be hard to do.
I always wonder about software ("game") they use. What is it ? Does everybody has some specific in-house developed thing ? Or ir there some off-the-shelf solution ? How does it compare to something like Asseto-Corsa. As i know it does take some shortcuts with its simulation.
Hi Dude! The Software that is used in the Video is rFactor Pro I instantly recognized it. It is basically rFactor 2 but more "enterprise-ish" with more specific features for these use cases and more telemetry output and things like that. So the next nearest thing would be rFactor 2 because they share the same physics engine from what I understand. Another software that is used by Mercedes for example is Panthera from Cruden.
Hi Scott, excellent video. I have a basic motion simulator from Pro Simu. It has 2 actuators in the front to simulate heave, roll, bumps ect and a traction loss actuator at the rear. It would be great to know how these 4000 euro sims stack up to the ultra pro ones you tested.
The limitation on testing the car on track is one of the stupidest ever. It forced teams to spend literally millions on building and running the simulators. Nice cost reductions there 😂
It was quite sensible before the cost cap, now i 100% agree with you - if you'd prefer to smash around silveratone for 3 days rather than use a sim you ahould be able to. Prior to the ban teams had whole 2nd teams for testing, extra cars, trucks, pit crews aletc. And Ferrari had (still have) their own track. Main issue now is engine allowance - i dont think many teams would want to add miles to their engines. But they could use a surrogate engine for aero testing. Bring it back!
It’s definitely nit-picky but the colour reproduction from the projectors seems awful. Assuming it would not add too much latency, replicating this with either a modular panel type screen or large curved displays with the ‘optical bezel removers’ would be very cool.
Hello Driver61; I have a serious question also😅; did they also develop the software, like the tracks, realism, the surroundings, ? Or do they use it from existing sims like AC? And if they make it themselves; why do they nail it so perfectly, and companiies like kunos or iracing cannot..? And why don’t they sell the software part also to those companies; so we all can buy it? It has been payed for already so they make double profit.. Again, just talking about physics and surroundings, not the motion hatdware etc
4 วันที่ผ่านมา +1
Exceptional performance is usually an exponential cost.
Interesting, as I want to move from my current sim, I would like to know if you need to pay a monthly subscription to play it and do you have to buy the cars you want to race with?
Can someone (with lot of money) could buy this for private home usage ? And does it work with normal game (would be fun to have a spin in this thing playing Mario Kart😅 )
Wouldn't it be interesting if one day simulators were a part of a teenager's driving training as they went through the process of learning to drive safely before getting their permits and driver's license. In 2022, 2,883 teenagers 13-19 years old died in car crashes in the USA. And in 2020, 227,000 teenagers were injured in motor vehicle crashes. Saving one life might be worth considering the prospects.
Just because he used some merc footage doesnt mean thats the answer. He also used Sauber. I think the most likely answer is McLaren, given they are a known customer of the company in question.
We used simulators extensively for aircraft testing. It is an invaluable tool for aircraft development. I’m surprised to hear of latency issues on F1 simulators as that has a relatively easy solution.
@@Zerobar78. I don't think you'd need latency as low for flying as you'd need to feel the car step out. You'd have tooons of warnings from the plane that it's gonna stall for example, whereas you're permanently on the limit with a race car. I think a big part of lowering the latency here is the hardware needed to make these quick adjustements too
The same latency is present in aircraft simulator rigs as driving rigs. you're not flying on the limit, trying to fly on a knife edge, unless you're in a flight sim doing dogfights or something similar.
The real skill scale: the price of your sim Fighter Jet (why ask?) F1 ($$$$$$$$$) Passenger Aircraft ($$$) Sanctioned Racing ($$) 24 Hours of LeMons Car (US$500) $/g is just so good at Lemons.
You don't need $2.5 million to build a sim for your home, and you can get a great sim experience with cheap hardware. Realistically you just need a Logitech G920, but $1.5K will get a nice entry sim, with $5-10K getting you an absolutely insane simulator with triple monitors, a real race car seat, an insane gaming PC, and a Simagic wheelbase, wheel and pedal set. But it's the SOFTWARE, not the HARDWARE, which is the most important part. These sims are basically video games and there's a bunch of them...iRacing is the most popular competitive sim and all the F1 drivers are there...just get a cheap setup and you will be 97% of the way there. The reason this setup in this video costs $2.5 million is because of the motion system (3 feet of movement across 6 axis). I cannot recommend iRacing more highly.
You need it to hold your head in the correct position (you could custom fit a new headrest, but throwing the helmet on is easier). It also correctly limits your vision to the same angles as when you're in the car, and allows you to talk with the sim engineers using the standard race comms systems.
The simulator platform knows where it is at all times. It knows this because it knows where it isn't. By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is Greater), it obtains a difference, or deviation. The motion control subsystem uses deviations to generate corrective commands to drive the platform from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is. Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't.
My sim consists of a Logitech G923 wheel and pedals, gear lever and simple desk and gaming seat that I have to secure to the wardrobe with a piano stool and a guitar speaker for it not to move (it still does a little when I push the brake). I suppose it could fall a little short of F1 teams' expectations...
you could still perfectly learn to drive, and if serious enough, win races of drift events in real life! try it
I bet this pro sim is still more quiet than g923 :D Mine sounds like a drill
i thought i was the only one who put my amp behind my chair😭
And you have fun! That’s 🎉🎉
@@docmain999griptape under a diningchair was my setup for years, although a cockpit and better wheelbase and all that sort of stuff adds alot to your quality of life.. it doesnt make you a quicker driver
My room mate blows next to me when there's wind and spits when it's raining. He also kicks me depending on the crash. Much more real phyisics than those "top" simulators.
Peak Immersive experience
What about snow? Or rallying in wet mud?
I hope you shower him with “champagne” after a win
Play rally games with him and see how he simulates mud
@ 😂
Wait. Is this the british TH-camr/ Racing driver that holds the Lap record at brands hatch???
Correct sir.
Why would you write racing driver and lap with capital letters, but British and Brands Hatch in lowercase?
@@jansz1589priorities, my Friend. also i think it Causes issues in the truly british. i kNow it buggEd me.
Still holding onto that one.... just.
how did you know that?! he never mentions it 🤣
The other team that he cannot mention is Cadillac. By the way
This was probably made before the Cadillac announcement
@unklesuga1644 thats why I am saying that. Cadillac knew for awhile they were gonna be 11th team. As soon as Micheal Andretti stepped down that was what F1 wanted.
Or Audi
@blade0667 i do actually forget about their team🤣 you're right. IF Ferrari is using that Simulator and Ferrari are supplying Cadillac's engines it makes sense
it has to be Haas
People seem perplexed that theres so much trust in simulators but NASA used various analog simulators for the Apollo missions in the 60's.
I m very perplexed because ALL teams missed the ‘porpoising’ effect and coming from a mech eng with 40 years exp. in CFD / STR analysis because the repeated bouncing up and down should be simple to solve I think !
Several of the Apollo astronauts were Marine or Navy pilots so they were already trained for high stress environments but of course still needed to practice working the specific controls for the Apollo flight plan so plenty of simulators of all kinds were created, truly amazing for 1960s and 70s 😎
@@philippemichelvidori7248 it's not really just bouncing up and down mate! It's a rather complex problem that I'm not exactly competent to explain to you having only a basic understanding, but what I can tell you is that fluid dynamics get complicated really quickly and when you have to consider flow on all sides, in the venturi tunnel, around the wing, etc. and on top of that distinguish between turbulent and laminar flow... it gets utterly mad.
@@philippemichelvidori7248 The problem with porpoising is that the windtunnel models are static and can't move up or down with regards to downforce, it's specifically to prevent the model from hitting the belt they run on as it would likely explode and create massive damage to both the wind tunnel and the model. So they're stuck relying on CFD simulations which are very performance intensive, you can either run a very accurate simulation on a small section of the car or run a more complete but inaccurate simulation on the whole car, or somewhere in between but never fully both. Porpoising is also a very complex feedback loop making it especially hard to simulate accurately even without the issues i mentioned before.
@philippemichelvidori7248 porpoising IS easy to resolve, you just run the car stiffer and higher.
But if you also want to be fast, you have a particularly complex problem.
You can't 1-1 simulate the g-forces the inner ear feels in a sustained way, but you can do it to some parts of the body. Eg your skin won't know the difference between being pressed into the side or your seat in a corner vs your seat pressing into your skin in a corner. Some sim rigs use this trick to simulate the sustained Gs with movable panels in the seat. Don't think it's 100%, but it does sound like 1 more level of immersion where you could quite literally be driving by the seat of your pants ;)
You can! Just make the sim rig actually move, and to sustain the g force, it should move miles away with wheels... Ah crap I just described actually driving the racecar
@@kerimca98 lol
A g-suit could simulate the sensations of g-force and the brain would interpret that in a similar way
@@fookapi that's with compressed air, right? not sure if that would have the response time they are looking for. but yeah, there are other ways to simulate Gs without moving the head about
@@alex8373 I’m imaging a full body suit that can apply pressure to different parts of the body to simulate forces. So unlike a traditional g-suit (which I wore as a military pilot), it shouldn’t just compress the legs. Just an idea and yes, it probably wouldn’t be fast enough.
Interesting video, but I would have liked it go way more into detail. “Simple” things like how they manage to keep latency low, make use of synchronization between visuals and movement of the platform, perhaps even sound (you didn’t mention that at all). But more importantly, you sidestepped *how* teams translate a part that only exists on the digital drawing boards to be “fitted” onto the virtual car and then influences the way the simulator behaves. To me, that is the most interesting aspect of this workflow.
Do these simulators have custom software for graphics and track specs? Did they laser scan all the tracks themselves and create their own sim software or are they using 3rd party data? Super curious!
We will never know
It's all straight from Forza Motorsport.
If I’m not mistaken some of the guys from Iracing started working on a project that was based off their models on Iracing
@ that’s what I’m wondering. Like if you’re 2mill in a sim. Is that “enough money spent, let’s use iRacing”. Or is it, “only our own personal laser scans of tracks meets the standards for our super expensive sim”
0:33 Kubica mentioned RAAAHHHH
better then my sim rig thats for sure
God could you imagine the grief you'd get from the wife? I have a hard enough time convincing mine to deal with my simple desk + chair setup
@@naplesbluesrt i do have a proper rig but i can't use it as i kept stubbing my toe when i got out of bed due to how close my desk is to my bed and just how big the rig is lol
@@The_Rally_Forge I covered the edges of my rig with neoprene pads lol
AWESOME! Great video guys :)
When you guys collabing?!
I agree Will
I bet Mazepin could find a way to make one of these spin around, or even on his best day crash and break the whole sim.
What is he doing nowadays?
@@Moonfaster fighting with the Russian army in Ukraine.
Maz once crashed a bike-hometrainer...
Funnily enough he was the only driver on the 2021 grid to not use a simulator
(Haas are broke)
you can replace this with Lance Stroll and you'll have a more accurate example
The other team is McLaren. The names of Ferrari and McLaren show up in the SEO results in the short description of SAE International's website when they covered Dynisma as a startup in 2022. I wonder if McLaren decided later to keep their supplier info private, while the search engines still have that information and the webpage is still up.
Best competent video about pro simulators available public.
That high-pass filter trick on the acceleration is really cool. So simple!
0:34 POLAND POWER! 💪💪🇵🇱🇵🇱
wruci
Sustained g-force is possbile. AMST Desdemona offers 3 g sustained - being a centrifuge. It probably can't achieve the necesary latencies but all I want to say is, you can simulate sustained g-forces.
I was waiting for someone like you to try a real F1 factory team simulator!! I’m so excited to watch it
Thank you for showing this. Has been a bit of a mystery as to how this all works and how they can be so reliant on it. Really cool to see this.
I wonder if you could simulate gforces in a big wheel with a rotating capsule, if you turn it a bit upside down and accelerate a little over 1G you could counter act the earths g force. The question would be how hard is it to increase Gs fast, can you go from 1 to 6 in fractions of a second. Maybe with a clutch or moving the capsule outward on tge wheel, or both.
This puts my moza setup to shame…. 😭 the only thing I have that they don’t is a fan on the base to simulate wind 😂.
My sim consists of a camping table as the desk, with my G29 wheel secured with a chopping block to keep it in place. For my pedals it’s under the camping table for blocks of wood so it doesn’t slide around on my carpet under braking. My gear shifter is on a container with every Harry Potter book staked on eachother as the stand with a flat wood plank and the top with the gear shifter secured to that. It also has wood counterweight blocks.
In theory easy to simulate g-forces. But the entire thing inside a centrifuge, like used for g simulation for fighter pilots.
That was my idea, then the car can rotate 360 ( if needed) to simulate g force. Then the car itself can be on rails itself and go up down the arm spinning, like a camera zoom steeper motors. To simulate another g force potential. While it can also increase / decrease g force as needed ( physics). While the main spinning motor can be slowed and speed up to also simulate more. The guy said you only need to get a percentage of the way there, so 2.5 g feels like 6 in sim ? 😮... I want to try it please !!
14:20 if your FFB is set correctly you can feel exactly what the rear is doing and how much weight is on each tyre in Assetto Corsa.
I was literally asking someone about this yesterday, and you go and make a video just for me to answer my curiosity. You are so generous! :)
bot
@danieldorn9989 bleep bloop
As a professional in simracing I've tested and installed many systems. The best of all is the one I still use since 12 years, it also have such frequency filtering no latency, it is affordable you can see on my channel. In combinaison with dlp projectors it has real 5ms latency from my hands to the feel and picture
I never feel that alive feeling in any other motion thing
Maintenance, replacement parts and the electric bill add up too….!
9:47 Or a human centrifuge with a swiveling basket and rapid spinup/spindown could do it. :)
Pressure suits could replicate the sensation. This tech is already developped
@@Supersonic_racing Pressure on the skin is one thing. Feeling acceleration in the inner ear is another important thing - which can't be influenced by pressure suits.
You are kind of right. It was already built for aerospace. See AMST Desdemona.
bang on
Really good video - explains all the fundamentals at play in the best motorsport simulators. 3 to 5 milliseconds latency - this is very good, but not the best available in F1 - agree totally that it is one of the critical things, but this sim is not the best for this particular metric. Am certain that they are working to improve this further. Driver reaction time is about 20 milliseconds or so, thus the latency is excellent compared to older driver in the loop simulators. Agree that the slight remaining latency is the main cause of drivers feeling sick - the world normally reacts faster in a real race car. Some really good and honest info about how the lower G-force is used to give a driver the most realistic feeling one can have. Loved it!
Fantastic video! Masterly created and on a very interesting topic at the heart of F1 development. My F1 sim consists of a MX5 and while it also does not achieve the same G-forces it does replicate without any delay many of the minute and sometimes major feelings like the rear end rotating mid corner and stepping out.
Does F1 or the FIA limit sim driving time? Or can teams and drivers just use the sim as much as they feel they need?
I always think that the best way a simulator could traduce g forces could be that the whole unit would be like a ball, where the pilot gets in, and inside of it , it would have like a cockpit, and with enough distance, rounded screens like the sphere in las vegas, but small size. Then, that ball-like unit would have the ability to rotate in any direction with a mechanic structure in the outside, but when youre sitting in the cockpit, with the seatbelts on, the horizon in the screens would be alingned properly in your field of view, so, when breaking, the ball viewed from the outside would rotate forward in the range of motion between 0 and 90 degrees depending on the strengh of the break, but viewed from the inside the pilot would have the screens with him so he wouldnt notice he is inclined towards the floor but he would only feel the weight of gravity as g forces on the break. And that same idea for the lateral g forces they feel when turning on a corner, or to feel speed G forces when accelerating inclining the ball backwards having the pilot facing the sky (without him noticing cause hes inside the screens sphere).
Instead of having a big screen attached to the floor and even if you have a simulated cockpit that moves and has some forward or lateral inclinations it wouldnt ever be able to incline much, without it loosing realism because the screen is separate from it and it would break the feeling of a real car-view behavior.
I got this idea from a f-16 or f-18 flight simulator I tried once, but it was different than this idea, it would be like 2 cockpits attached in front of the other, then you would have like an axis between them, so youwould get into the cockpit and you would only see like an inside of an airplane cockpit and like a screen where you would see like the airplane view or something like that. And then it would "take off" and you would begin to feel the G force of a real f-16 but viewed from the outside, the 2 cockpits would be spinning like a fan or a blender, and the cockpit would be on each blenders blade looking towards the axis. That way you have the g forces only in one direction, you would only be able to feel "speed" G forces. Not breaking or lateral G forces. Thats why I thought about the other Idea. I dont know much about simulators but it has always been interesting to me that Ive never seen one with the same idea I have in mind, not even in f1 simulators. Maybe thats a bad idea but it makes so much sense to me that it would be way more realistic for the pilot.
This! I agree totaly. Something similar is in fun parks where you go inside "simulator" for a rollercoaster ride or similar
You'd still be limited to roughly 1G, much less than the forces in a real car, and you might get some unwanted effects or cues to the driver due to having to rotate forwards, backwards or sideways to apply the g forces
Not able to apply high enough forces, will move the pilot in a manner inconsistent with the vehicle, and a hugely complex video problem (to create something less real).
great video. i really wish home sims like iracing would allow us to render only the elements that we need like in 1:16. if i could only render the outside world and hide the dash i would be a very happy camper
I've long been curious about F1 teams' simulators, so this was really great.
I wonder if parallelization could be used to pre-compute all the predicted values that can happen on the next frame or frames... I mean, if the value of something is '5' in this frame it won't be 50 000 the next frame. It will be 6 or 8. Add 16k predictions matrix and you can pick the solution from a storage. Kind of what our brains do all the time... Computing something takes time but if we can solve all possible predictions of what is going to happen next then you only need to execute.
Anyone knows in wich engine that sim is based??
(rFactor pro maybe?)
An active harness and G seat helps with feeling sustained forces.
Qubic belt tensioner!!
@@michaelhuyer1479 Yep, have one. 😁
@@seattime4075 next year i will get one for my V3 from motion system EU👍
Which software are using? Assets Corsa or something native?
I'm a bit confused. For the inner ear breaking and pitching forward feels exactly the same. Can't you just simulate g forces by pitching/rolling? Sure you can only simulate 1g of cornering/braking acceleration, but this could be compensatet for with additional weight the driver wears. Instead of the huge screen you would obviously need vr googles or somthing comparabel to fool the driver into experincing only the g forces without a relation to their surrounding. The only downside would be the constant weight pulling driver down. But this way you could have sustained g forces throughout longer corners/braking sections. This exact same principal is used for flight simulators. Maybe I'm just overseeing something, but this would be a logical way for me simulate sustained g forces. Maybe you can help me see my flaw in logic here?
The inner ear knows nothing about the additional weight on the body. For the ear, 1 g is still the maximum sustained force by pitching - leaving 0 g felt towards the butt. If you want sustained g-forces even above 1g that the ear can sense, you need a centrifuge (see AMST Desdemona for example).
0:20 Apex GP ;)?
he is not saying no so you might be on to something
No.
I really enjoy being educated about these subjects. Thank You!
mmh, i felt bad spening 800$ on my PC.
The consumer/console games don’t do a good job of decoupling the front and rear axles, but I’d imagine from the remarks about the feel of the tear end under braking, these are beyond an order of magnitude better in that regard. So cool. Would like to know if this company is supplying the NASCAR teams as well.
Spa looks exaclty like assetto corsa's model. Are they using a modded version of the game? Or just the track scan?
Really nice video! So interesting to have a focus on such a topic, with a good mix of facts, interviews and visuals. Keep it up!
Hey scott can i ask you something? is the spanish track an AI? because its sound just like your voice, if is a real person well congrats because its so good
I kinda expected the bigger rigs to use some sort of 6 axis system to use gravity to simulate G-force, though i suppose when the cars can pull 6G maybe that is less useful.
Porpoising must have been like the day of reckoning for simulation departments on every F1 team
Interesting that platform pitch and roll aren't used as surrogates for acceleration as they are in flight simulators on Stewart platforms. One reason is because the visual is not on the platform and would have little screen space left to compensate for the tilt, particularly in roll, and another is that these types of cues produce undesirable rotation rates in the inner ear, ever so more with the required onset of the acceleration, leading to motion sickness. With flight sims, braking is the worst offender, and difficult to tune the filters to achieving a compromise. Surprised 3ms latency (and they probably meant throughput) is required, as the human animal has a mean 1/3 sec time constant - I would have thought 30ms would suffice in these high gain applications. Nice work.
Is the cost of the simulator included in the team's cost cap? If a team cannot afford the best simulator, is that an unfair advantage?
Why "unfair"? If they can't afford the best race engineers, is that also "unfair"?
This was a brilliant watch, your work is incredibly informative and well put together. Thank you Scott.
Is audio fed to the driver, engine noise, vibration etc?
Does anyone know if the Iracing guys took their model to develop software for the race team sims. I can’t remember where I heard that. So a lot of these are based off the original Iracing models
So they derive the acceleration once more to get jolt, that is the rate of change of acceleration. Using a tilting sim would allow for simulating a weak but rather constant centripetal force, but it would produce some other unwanted forces too. It's better suited for aviation sims, where constant g-loading builds up rather slowly, so the required reaction speed is much slower, apart from the little bumps you'd feel from aerodynamic turbulence.
The rate of change of acceleration is Jerk, not jolt.
And the rate of change of jerk is snap, then crackle and then pop. Don't say physicists don't have a sense of humour
@@PaulHickford Whatever order of magnitude we look at, at some point we reach a degree of amalgamation where bones behave like play-doh, planets behave like liquids and space-time behaves like a fabric of information. Isn't the universe fascinating!
- some ER surgeon pondering about his last job, probably
I'd love to know what software it runs.
I found the part where the gentleman explained the “high pass filter” interesting. They use the “jerk” (technical term) to impart the feeling of changes in acceleration and deceleration. The drivers’ imagination does the rest! Cool.
I think they could add wires to the helmet to drag your head around to simulate g=force.
I like the way you think
@@stevenhaggar132 Been giving it a bit of thought lately. I reckon three wires, two diagonally out the front and one going straight back. I would mount that to a bearing mounted ring so you can turn your head and mount the ring to a helmet with velcro. I figure you want a system that you can break free off in case of a fault in the system, no responsibility taken for broken necks or whiplash... :)
Edit: Scratch the whole ring and bearing idea. Just mount the wires to the top and centre of the helmet, that should allow enough freedom to turn.
I like it
@@stevenhaggar132 Thanks for the vote of confidence, I emailed Dynisma and passed on the idea about 10 minutes ago.. Not sure if we'll ever see this for consumer grade simulators but if you are spending mega bucks it wouldn't be hard to do.
If only Nick Wirth had a simulator this good back in 2009, with his "CFD only" built F1 car
I always wonder about software ("game") they use. What is it ? Does everybody has some specific in-house developed thing ? Or ir there some off-the-shelf solution ?
How does it compare to something like Asseto-Corsa. As i know it does take some shortcuts with its simulation.
Hi Dude! The Software that is used in the Video is rFactor Pro I instantly recognized it. It is basically rFactor 2 but more "enterprise-ish" with more specific features for these use cases and more telemetry output and things like that. So the next nearest thing would be rFactor 2 because they share the same physics engine from what I understand. Another software that is used by Mercedes for example is Panthera from Cruden.
What are they using a self-made sim ?, or is it something like ac ?
Rumor is that it is Rfactor Pro, the one unavailable to the public
Hi Scott, excellent video. I have a basic motion simulator from Pro Simu. It has 2 actuators in the front to simulate heave, roll, bumps ect and a traction loss actuator at the rear. It would be great to know how these 4000 euro sims stack up to the ultra pro ones you tested.
I have aTraction loss and sway Motion Sim,i drive in VR.Its fantastic with drift cars around corners,super real Feeling,my Sim cost about 15000 Euro
0:35 GIGAKUBICAAAAA \o/
Great content, appreciate it!
one of the very few video's that I actually liked.
The limitation on testing the car on track is one of the stupidest ever. It forced teams to spend literally millions on building and running the simulators. Nice cost reductions there 😂
It was quite sensible before the cost cap, now i 100% agree with you - if you'd prefer to smash around silveratone for 3 days rather than use a sim you ahould be able to.
Prior to the ban teams had whole 2nd teams for testing, extra cars, trucks, pit crews aletc. And Ferrari had (still have) their own track.
Main issue now is engine allowance - i dont think many teams would want to add miles to their engines. But they could use a surrogate engine for aero testing.
Bring it back!
Great interesting video...do other race series teams use these as well.. like IMSA/WEC endurance racing teams??
Yes. Peugeot just put out a video of Theo Pourchaire in their simulator the other day. Even F2/F3 teams have simulators (though not as advanced).
Which software do they use?!
a part from Pete Windsor, nobody can touch Scott's content quality on motorsports on yt
It’s definitely nit-picky but the colour reproduction from the projectors seems awful. Assuming it would not add too much latency, replicating this with either a modular panel type screen or large curved displays with the ‘optical bezel removers’ would be very cool.
yeah, that's what I noticed too. All that money just for the most Washed Out image you've ever seen.😂
Hello Driver61; I have a serious question also😅; did they also develop the software, like the tracks, realism, the surroundings, ? Or do they use it from existing sims like AC? And if they make it themselves; why do they nail it so perfectly, and companiies like kunos or iracing cannot..? And why don’t they sell the software part also to those companies; so we all can buy it? It has been payed for already so they make double profit.. Again, just talking about physics and surroundings, not the motion hatdware etc
Exceptional performance is usually an exponential cost.
Just let AI Drive, don't need to pay millions to real drivers 😂😂😂😂😂
Actually... They already have this... 😅 It's first year was a disaster but it's going to get fun... Would be cool to see AI vs. Human.
Where do I get one of these sims? They look fun.
Interesting, as I want to move from my current sim, I would like to know if you need to pay a monthly subscription to play it and do you have to buy the cars you want to race with?
Can someone (with lot of money) could buy this for private home usage ? And does it work with normal game (would be fun to have a spin in this thing playing Mario Kart😅 )
2:27
Hamilton in Singapore 2018, "Hold my tea"
The day they invent a super fan in which they can manipulate the density of the air, it will be possible to simulate G-force in racing simulators.
And what Sim are they using?
EA Sports F1 24😅
I would love to know what software they use. Or is that secret?
Wouldn't it be interesting if one day simulators were a part of a teenager's driving training as they went through the process of learning to drive safely before getting their permits and driver's license. In 2022, 2,883 teenagers 13-19 years old died in car crashes in the USA. And in 2020, 227,000 teenagers were injured in motor vehicle crashes. Saving one life might be worth considering the prospects.
Virtual Spa versus the Real one? Cool concept to test on any track.
The Hidden Team is Mercedes (water marks top left)
Just because he used some merc footage doesnt mean thats the answer. He also used Sauber.
I think the most likely answer is McLaren, given they are a known customer of the company in question.
would be interesting to know about their software they use
Linus needs to see this place. I wanna see the computer specs. 192 core threadripper or 9800X3D? RTX 4090 or RTX A100? :D
"I changed my racing sim. AGAIN!"
We used simulators extensively for aircraft testing. It is an invaluable tool for aircraft development. I’m surprised to hear of latency issues on F1 simulators as that has a relatively easy solution.
What's the solution, and why haven't they applied it to what you saw in the video?
@@procatprocat9647 the answer is probably money.
@@Zerobar78. That's probably it. IIRC, unlike a racer, you can't do your first test flight until you spend hundreds of hours in a simulator.
@@Zerobar78. I don't think you'd need latency as low for flying as you'd need to feel the car step out. You'd have tooons of warnings from the plane that it's gonna stall for example, whereas you're permanently on the limit with a race car. I think a big part of lowering the latency here is the hardware needed to make these quick adjustements too
The same latency is present in aircraft simulator rigs as driving rigs.
you're not flying on the limit, trying to fly on a knife edge, unless you're in a flight sim doing dogfights or something similar.
Why is track testing restricted?
$$$
So did the yellow platform on the thumbnail make a difference?
Awesome video!
Are they using projectors for the screens? Wouldn't that introduce very noticeable input latency?
No and no.
ok everything is good, but why in reality often happen that the things dveloped with simulators go very very badly on tracks?
The real skill scale: the price of your sim
Fighter Jet (why ask?)
F1 ($$$$$$$$$)
Passenger Aircraft ($$$)
Sanctioned Racing ($$)
24 Hours of LeMons Car (US$500)
$/g is just so good at Lemons.
2nd f1 team use this supplier simulator ( used for 2025 car) is McLaren F1 team as per my knowledge. i guess.
If I had the money, I'd buy one.
Wow. That is amazing tech. I need one
Somewhat makes on track testing restrictions a bit redundant.
I drive right past this company everyday to work! Never knew what they did. Its in a tiny village in the middle of nowhere.
Why did the teams have porpoising issues?
Many tiny reasons.... Like track surface texture makes a difference, cos of the way it affects air flow etc
You don't need $2.5 million to build a sim for your home, and you can get a great sim experience with cheap hardware. Realistically you just need a Logitech G920, but $1.5K will get a nice entry sim, with $5-10K getting you an absolutely insane simulator with triple monitors, a real race car seat, an insane gaming PC, and a Simagic wheelbase, wheel and pedal set. But it's the SOFTWARE, not the HARDWARE, which is the most important part. These sims are basically video games and there's a bunch of them...iRacing is the most popular competitive sim and all the F1 drivers are there...just get a cheap setup and you will be 97% of the way there. The reason this setup in this video costs $2.5 million is because of the motion system (3 feet of movement across 6 axis). I cannot recommend iRacing more highly.
I mean with limited real world testing the confidence is kind of relative. You actually have to trust it at some point.
I think I know what I want for Christmas...
Whats with the helmets in the sim?
You need it to hold your head in the correct position (you could custom fit a new headrest, but throwing the helmet on is easier). It also correctly limits your vision to the same angles as when you're in the car, and allows you to talk with the sim engineers using the standard race comms systems.
@@alan_davis Thanks for that.
The simulator platform knows where it is at all times. It knows this because it knows where it isn't.
By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is Greater), it obtains a difference, or deviation.
The motion control subsystem uses deviations to generate corrective commands to drive the platform from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is.
Consequently, the position where it is, is now the position that it wasn't, and it follows that the position that it was, is now the position that it isn't.